Suicide Risk and Risk Factors After Hospitalization for Depression

Aaltonen, K., Sund, R., Hakulinen, C., Pirkola, S., & Isometsä, E. (2024). Variations in suicide risk and risk factors after hospitalization for depression in Finland, 1996-2017. JAMA psychiatry, 81(5), 506-515.

Key Points

• The risk of suicide among patients hospitalized for depression is extremely high in the first days after discharge, with incidence rates of 6,062 per 100,000 person-years on days 0-3 and 3,884 per 100,000 person-years on days 4-7 after discharge.
• Several factors significantly increased the risk of suicide in the short-term after discharge, including severity of depression, current suicide attempt, previous suicide attempt, male sex, older age, and higher household income.
• The relative risk associated with different factors showed varying temporal patterns over the 2-year follow-up period: some remained constant (e.g. male sex), some decreased over time (e.g. current suicide attempt), and some increased over time (e.g. alcohol use disorder).
• While providing valuable insights into imminent suicide risk factors, the study was limited by its reliance on register-based data, which lacks detailed clinical information about patients’ status at discharge.
• Understanding the dynamic nature of suicide risk factors over time is crucial for developing targeted interventions and improving care for high-risk patients with depression after psychiatric hospitalization.

Rationale

Depression is associated with the highest risk of suicide among mental disorders, with up to 7-8% of men and 4% of women with severe depression dying by suicide (Nordentoft et al., 2011).

The risk of suicide is particularly high in the period immediately following discharge from psychiatric hospitalization (Chung et al., 2019).

However, most previous studies on suicide risk factors have focused on long-term follow-up periods and modeled risk as constant over time (Franklin et al., 2017).

This approach contrasts with the current understanding of suicidal behavior as influenced by volatile state-dependent factors and more stable trait-like predispositions (Turecki & Brent, 2016; van Heeringen & Mann, 2014).

There is a critical need to examine suicide risk and risk factors in the short-term period after discharge, particularly for patients hospitalized for depression.

Additionally, it is important to investigate whether the relative importance of different risk factors changes over time following discharge.

This study aimed to address these knowledge gaps by examining suicide incidence and risk factors starting from the first day after discharge up to 2 years, using comprehensive national register data from Finland.

Method

This was a population-based cohort study using Finnish national registers, including hospital discharge, population, and cause of death registers.

Procedure

The study identified all psychiatric hospitalizations for patients 18 years and older with depression as the principal diagnosis from 1996 to 2017.

Patients were followed for up to 2 years from their last discharge, with each discharge marking the beginning of a new follow-up period.

Sample

The study included 193,197 hospitalizations among 91,161 individuals. Of these individuals, 56.2% were female, and the mean age was 44.0 years (SD = 17.3).

Measures

Data were collected from national registers on:

• Psychiatric diagnoses (ICD-10 codes): The study used the International Classification of Diseases, 10th Revision (ICD-10) codes to identify and categorize psychiatric diagnoses.
• Global Assessment Scale (GAS): Used to assess the overall severity of symptoms and functioning at admission.
• Suicide attempts: Data on suicide attempts were collected from the Care Register for Health Care using ICD-10 codes X60-X84, Y87.0, and Z91.5.
• Sociodemographic factors: E.g. age, sex, living situation, income, education. These were retrieved from the Population Data of Statistics Finland, which is annually updated.
• Cause of death: This information was obtained from the Causes of Death register maintained by Statistics Finland. Suicides were identified using ICD-10 codes X60-X84, Y87.0, and Z91.5.

Statistical measures

The study used:

• Poisson regression to estimate suicide incidence rates and incidence rate ratios (IRRs)
• Hazard functions and hazard ratios estimated using penalized cubic splines
• Cumulative incidence of suicide estimated using the Aalen-Johansen estimator

Results

Incidence of suicide after discharge:

• Extremely high in the first days: 6,062 per 100,000 person-years on days 0-3
• Declined steeply but remained high: 3,884 per 100,000 on days 4-7, 2,474 per 100,000 on days 8-30

Short-term risk factors (0-3 days after discharge):

In the context of this study, incidence rate ratio (IRR) is a statistical measure used to compare the incidence rates of suicide between an exposed group to the incidence rate in an unexposed (or reference) group. 

For example: An IRR of 1 indicates no difference in risk between the groups, greater than 1 indicates an increased risk in the exposed group, and less than 1 indicates a decreased risk in the exposed group.

• Severe depression: IRR 2.18 (95% CI 1.37-3.52)
• Psychotic depression: IRR 2.36 (95% CI 1.35-4.07)
• Severe symptoms/impairment (GAS 0-19): IRR 5.26 (95% CI 1.46-15.39)
• Current suicide attempt: IRR 3.66 (95% CI 2.09-6.02)
• Previous suicide attempt: IRR 2.09 (95% CI 1.28-3.29)
• Male sex: IRR 1.96 (95% CI 1.31-2.96)
• Age 40-65 years: IRR 2.76 (95% CI 1.68-4.75)
• Highest income tertile: IRR 1.99 (95% CI 1.04-3.67)

Temporal patterns of relative risk over 2 years:

• Constant: Male sex, previous suicide attempt
• Decreasing: Severe/psychotic depression, current suicide attempt, lower GAS score, older age
• Increasing: Alcohol use disorder, substance use disorder, living alone, involuntary admission

Insight

This study provides crucial insights into the dynamic nature of suicide risk following psychiatric hospitalization for depression.

The extremely high risk in the immediate post-discharge period underscores the critical importance of intensive follow-up care and support during this time.

The identification of specific short-term risk factors, such as severe depression, current suicide attempts, and older age, can help clinicians prioritize resources and interventions for the most vulnerable patients.

The varying temporal patterns of risk factors over the 2-year follow-up period is a novel and important finding. It suggests that the relative importance of different risk factors changes over time, which has significant implications for ongoing risk assessment and management.

For example, while clinical factors like depression severity and current suicide attempts have the strongest association with risk immediately after discharge, their relative importance decreases over time. In contrast, factors like alcohol and substance use disorders become increasingly important predictors of suicide risk in the longer term.

These findings extend previous research by providing a more nuanced understanding of how suicide risk evolves over time after psychiatric hospitalization.

They highlight the need for dynamic risk assessment models that account for changes in the relative importance of different factors over time.

Further research could focus on:

1. Developing and validating predictive models for short-term suicide risk that incorporate these dynamic risk patterns.
2. Investigating the effectiveness of targeted interventions based on these temporal risk patterns.
3. Exploring the underlying mechanisms that explain why some risk factors increase in importance over time while others decrease.

Strengths

1. Large, nationwide sample covering all hospitalizations for depression over a 22-year period
2. Long follow-up period of up to 2 years
3. Use of high-quality national registers with comprehensive coverage
4. Examination of both short-term and long-term risk factors
5. Novel analysis of temporal patterns in relative risk over time

Limitations

1. Reliance on register-based data, which lacks detailed clinical information about patients’ status at discharge
2. Inability to capture changes in clinical status or risk factors during the follow-up period
3. Potential underdiagnosis of certain conditions (e.g., psychotic depression, alcohol use disorder) in clinical settings
4. Lack of information on outpatient treatment or interventions received after discharge
5. Potential misclassification of some deaths on the day of discharge

Clinical Implications

The results of this study have significant implications for clinical practice and suicide prevention efforts:

1. Immediate post-discharge period: The extremely high suicide risk in the first days after discharge highlights the critical need for intensive follow-up care and support during this period. Clinicians should ensure that patients have a clear safety plan and rapid access to mental health services upon discharge.
2. Risk assessment: The identified short-term risk factors (e.g., severe depression, current suicide attempt, older age) should be carefully evaluated before discharge and used to guide the intensity of follow-up care.
3. Dynamic risk monitoring: Given the changing patterns of risk factors over time, ongoing risk assessment should be conducted regularly after discharge, with attention to factors that become more important over time (e.g., alcohol and substance use disorders).
4. Targeted interventions: Prevention efforts should be tailored to address the most relevant risk factors at different time points. For example, interventions focused on managing depression symptoms may be most critical immediately after discharge, while addressing substance use issues may become increasingly important over time.
5. Resource allocation: Mental health services can use these findings to guide the allocation of resources, ensuring that high-risk patients receive appropriate levels of care at different stages following discharge.
6. Policy implications: The study supports the need for policies that ensure continuity of care and rapid access to mental health services for patients discharged from psychiatric hospitalization.

The significance of these results is substantial, as they provide a more nuanced understanding of suicide risk that can inform more effective prevention strategies.

By recognizing the dynamic nature of risk factors, clinicians and health systems can develop more personalized and adaptive approaches to suicide prevention for patients with depression.

References

Primary reference

Aaltonen, K., Sund, R., Hakulinen, C., Pirkola, S., & Isometsä, E. (2024). Variations in suicide risk and risk factors after hospitalization for depression in Finland, 1996-2017. JAMA psychiatry, 81(5), 506-515.

Other references

Chung, D. T., Ryan, C. J., Hadzi-Pavlovic, D., Singh, S. P., Stanton, C., & Large, M. M. (2017). Suicide rates after discharge from psychiatric facilities: a systematic review and meta-analysis. JAMA Psychiatry, 74(7), 694-702.

Franklin, J. C., Ribeiro, J. D., Fox, K. R., Bentley, K. H., Kleiman, E. M., Huang, X., … & Nock, M. K. (2017). Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research. Psychological Bulletin, 143(2), 187-232.

Nordentoft, M., Mortensen, P. B., & Pedersen, C. B. (2011). Absolute risk of suicide after first hospital contact in mental disorder. Archives of General Psychiatry, 68(10), 1058-1064.

Turecki, G., & Brent, D. A. (2016). Suicide and suicidal behaviour. The Lancet, 387(10024), 1227-1239.

van Heeringen, K., & Mann, J. J. (2014). The neurobiology of suicide. The Lancet Psychiatry, 1(1), 63-72.

Keep Learning

Socratic questions for a college class to discuss this paper

1. How might the extremely high suicide risk in the immediate post-discharge period inform changes in current discharge planning and follow-up care practices?
2. What ethical considerations arise when using these findings to assess individual patients’ suicide risk? How can we balance the need for targeted interventions with concerns about potential stigmatization or overly restrictive practices?
3. How might the temporal patterns of risk factors identified in this study challenge our current understanding of suicidal behavior? What new research questions do these findings raise?
4. Given the limitations of register-based data, what additional types of studies might complement these findings to provide a more comprehensive understanding of suicide risk after psychiatric hospitalization?
5. How might sociocultural factors influence the generalizability of these findings to other countries or healthcare systems? What additional factors should be considered when applying these results in different contexts?
6. How could the dynamic nature of risk factors identified in this study be incorporated into clinical decision-making tools or risk assessment protocols? What challenges might arise in implementing such dynamic risk assessment in real-world clinical settings?
7. Considering the finding that higher income was associated with increased short-term suicide risk, how might this challenge our assumptions about the relationship between socioeconomic status and mental health outcomes? What theories could explain this counterintuitive result?
8. How might the increasing importance of alcohol and substance use disorders over time inform long-term treatment strategies for patients with depression? What implications does this have for integrated mental health and substance use treatment programs?
9. Given the high risk of suicide in the post-discharge period, what innovative interventions or technologies could be developed to provide support and monitoring during this critical time?
10. How might these findings inform public health policies and resource allocation for suicide prevention at a population level? What are the potential economic and social benefits of implementing more targeted, time-sensitive interventions based on these results?